My AGD analog detector work

[ivconic]

Juste registered and checked with one sample board.
Huh! Shipping methods are fewer than what JLPcb offers and much more expensive!
10 times more expensive shipping from Ger to Ser... then from Chi to Ser, crazy!
12 boards ... the price is 4 times more then at JLPcb!
Overall... UI at their web site is not that friendly and easy as at JLPcb.
Naaa... I will stay with JLPcb.

[GeoMax]

Hi

AISLER, is not a German company. Its from the Netherlands. But only about 1000 meter away from the German/Netherlands boarder. I notice that several PCB companies in Europe takes "european prices" for PCB's what they send to China for manufacturing. Happend to me... My last PCB's I ordered at JLPcb in China.... Good quality, fast shipping to Germany and acceptable pricing.

[ivconic]

Hi

AISLER, is not a German company. Its from the Netherlands. But only about 1000 meter away from the German/Netherlands boarder. I notice that several PCB companies in Europe takes "european prices" for PCB's what they send to China for manufacturing. Happend to me... My last PCB's I ordered at JLPcb in China.... Good quality, fast shipping to Germany and acceptable pricing.

Yeah! Here too. In Serbia there are several "companies" (motherf...ing as..oles) that advertize themselves as "PCB Manufacturing Company".
First time I wanted to make "pro" double sided pcbs with metalized holes; I called one of those.
They asked for gerbers. 2 Days later a "beetch" called me back, informing me that it will take 25 days they to finish the pcbs and asked me to pay in advance 380 euros for 10 small pcbs!!!
In the very same moment while I was at phone with her; I was also checking JLPcb online and the price was 9.80 euros also for the same 10 small pcbs + 16 euros for shipping.
Affectionately, I don't know what I didn't tell her! Dirty, smelly, ugly...

[Auto-Mation-Assist]

I was able to update 6 of the AGD23-3 diagram pages and anticipate making some minor changes to the timing drawing to add some level shifter parts to outputs TS-3 and TS-4 even though they are not used. The supplied diagrams are for the AGD23-3 version for which the required receive board has already been laid out. Boards will likely be ordered before the end of this month. The parts that I have been waiting for are now expected to be shipped to me on about Mar 14th instead of in May.

I have attached a picture of a board built about a year ago (22-1) and the 23-3 version will be very similar with on relocations of mounting holes. Some parts get moved slightly.

[Auto-Mation-Assist]

Hi

AISLER, is not a German company. Its from the Netherlands. But only about 1000 meter away from the German/Netherlands boarder. I notice that several PCB companies in Europe takes "european prices" for PCB's what they send to China for manufacturing. Happend to me... My last PCB's I ordered at JLPcb in China.... Good quality, fast shipping to Germany and acceptable pricing.

The Co-Owner's office is the Netherlands while manufacturing is done in Germany. All my orders come directly form Germany. The AGD23 receive board is a four layer board and they do an excellent job. I like the fact that they are in part in the Netherlands since I was born and grew up there.

[Auto-Mation-Assist]

I have been been working on diagrams and also a bit more development work. I wanted to reduce the TX pulse width to 50us max, decrease my receive window from about 3.5us to 2us for my time slot 0, and on my TX board change the 20K 1W resistor to 9.1K to get close to the desired resistance of 628.7 ohms for the five parallel resistors. 9.1K should give 628.4 ohms,I would prefer 9.2K which would give 628.4 ohms. Ended but using 10K 1W since that was on hand for 632.3 ohms. And did some testing and the result were all good.

I attached a picture of the receive wave form that show the 2us RX window that the detector uses now. If you look at the waveform you can see that the detector normalizes its zero volt reference point on the center line. This picture is at the output of the front end variable amplifier. As you can see most of the waveform is below the zero line and is as designed and moves up and down depending on timing settings and environment. This picture was with a Coiltek Elite 9 inch coil.



I was using a older version of the AGD23 to do some verification work before populating one of the newer PCB's that I received a few days ago.



It gets to be pretty messy at times.

I have attached the diagram for my front end. I had calculated that it was going the require about 88dB of overall gain for the received board and to be divided between all its stages . The actual overall gain that I ended bu with 94.3dB or a voltage gain of about 25111 times for all the stages combined to the VCO input.

The below picture shows excellent frequency response of the first stage.

[ivconic]

Neat and clean work.
Congratulations on that.
You also have good measuring equipment in the workshop.
I also looked at the schematics you provided.
What is the final goal with this project of yours?
And do you have a video of how it works?
Sorry if I'm asking something you've already answered; I tried to read everything, maybe I missed it.

[Auto-Mation-Assist]

My has been my intent to design and build a analog metal detector that is sensitive enough to detect small gold flakes at a reasonable distance for my own use. Over the last several years I have refined the AGD23 design and very happy with its performance. I'm posting the diagrams with the hope that some of the information there will be useful to other persons who may want to build their own versions or adapt certain portions into their own efforts. Building one of these from scratch is a major effort and the complexity that is normally hidden in digital code adds to parts count. The receive board has about 374 parts that need to be installed and takes most of a day to populate one of my boards.

I plan on updating the three that I have built for our own use and call this project complete. Summer is getting close and its time to have some fun swinging them around with family and friends. In the future if time allows I might work on developing a interface for digital signal processing and timing for this detector. The receive front end and transmit board are perfectly suitable for doing that and would not require changes. If anyone is interested in doing that I would be happy to assist.

The front end design goal has been to utilize only linear parts to keep intermodulation (IMD) generation to a minimum and to allow fast recovery from overload due to transmit pulsing and its bad side effects. As a result you will not find and switches before the first amplification stage since those can act like a none linear device while switching and increase susceptibility to interference. The closets to this is the diode in series with Mosfet driver and it gets reversed biased almost immediately upon completion of the TX pulse by the coils kick back. Intermodulation distortion occurs when two or more signals are received by a system that has components that have a non-linear area when switching on and off just like a mixer in a radio receiver. Some times using a simpler design has benefits over a more complex one.

The drawing of the AGD23 front end that I posted in .pdf format is actually page two of the design. Page 3 (Signal filtering and sample and hold), and page 4 (Signal filtering and amplification) are still being redrawn to add component values

The 125kHz square wave test picture that I posted is a usual test done in audio design work to verify that there is no or little overshoot in the waveform and indicates its usability to ten times the test frequency or 1.25Mhz. Careful PCB layout is important here. Back in my prior years I built a lot equipment and had over 110 pieces of personal test equipment and this test was used a lot when gain flatness was very important.

I do not plan on making any videos.

[ivconic]

Thanks for reply!

[MartinB]

Impressive. I take my hat off to you.

[Auto-Mation-Assist]

I have been looking at the present version of the receive board that I use AGD23.2.A and my work on the updates to the potential future version AGD23.3.A. There is not much difference between the two right now. Two capacitors on the 7556 timer control inputs get removed since the 7556 version does not need them. Also three resistors get added and some values changed for the TS-3 and TS-4 timer outputs to make them compatible for driving 3 volt logic. These outputs are not used but handy for making some timing measurements. Also add is 750 ohm resistor between DTS connector and U12.1 pin 1 output. This compensates for a bit of signal loss there ( about -0.7dB) when adjusted to minimum gain. This is not important but I would prefer 0dB or close to it and thus the 750 ohm resistor addition is proposed on page 5 of the diagrams. I think I would also add a test point on the master timer output U6.1 Pin 5 on page 1 of the diagrams as a time reference trigger point.. Reference diagrams in attached AGD23.2.A-Updated.pdf file it contains all nine diagrams for the AGD23.2.A pcb..

I have been busy populating an additional PCB today and it took about 6 hours to pull the surface mounted parts needed and install them. All I have to do now is install the through hole parts, connectors, test points and several trim pots for VCO and indicator driver. The below picture of a AGD23.2.A receive board with just its SMD parts mounted. The very small ultra high speed dual diode array marked DA1 at the bottom left, and just below the 1 watt resistor marked 223 is the most difficult to install and very important to protect the ADA4637-1BRZ FET input OpAmps. It must show continuity in both polarities from any pin 2 of any of the three 1 watt resistors to ground. Refer to the attached diagram page 2 for the DA1 dual diode array and the three parallel 1W resistors.

Just the SMD parts mounted on board. Cell phone camera makes the board look canted.




I need to add 1 watt notations on the diagrams for R5, R134, R138, R144 (2512 size). Other other resistors are 1/4 watt (1206 size). Exception is timing polarity jumpers which are zero ohm 805 size. Most caps are also 1206 size. Alum caps vary in size.

[ivconic]

Very neat and beautiful.
But without seeing how it works, for even a few seconds; it will remain only text and image on "paper".
You said you wouldn't do video demonstrations, ok, I respect that.
But do you agree that all this posting of pictures and stories of yours is ... a bit unnecessary?
Publishing the schematic is actually both important and useful, thank you for that!
But the schematic is too complex for someone to just decide to build it... at least until see that thing in action.
A few seconds of video of it reacting to a target... won't kill you.
There are tons of detector videos. Most don't show much.
But it is possible to "feel" and "see" at least part of the performance through the eyes, at least those who have practice and know how to evaluate can do that.
Or is it just not working right and you're embarrassed to show it?
Images with oscillograms are not indicative.

[Auto-Mation-Assist]

Very neat and beautiful.
But without seeing how it works, for even a few seconds; it will remain only text and image on "paper".
You said you wouldn't do video demonstrations, ok, I respect that.
But do you agree that all this posting of pictures and stories of yours is ... a bit unnecessary?
Publishing the schematic is actually both important and useful, thank you for that!
But the schematic is too complex for someone to just decide to build it... at least until see that thing in action.
A few seconds of video of it reacting to a target... won't kill you.
There are tons of detector videos. Most don't show much.
But it is possible to "feel" and "see" at least part of the performance through the eyes, at least those who have practice and know how to evaluate can do that.
Or is it just not working right and you're embarrassed to show it?
Images with oscillograms are not indicative.

I felt like I was being lectured for the second time after reading this message.

And then this: "Or is it just not working right and you're embarrassed to show it? "

Your using the wrong strategy to attempt to belittle me to get your way. First there is nothing to be embarrassed about and I'm to old at the age of 76 to play games. I'm only here to share information about my project with persons interested in building it or something like it. If my project or information is not suitable for you then I'm sure that you can find something you like since there are lots of good things on the geotech1 forum to choose from. If you wish to post something that helps me improve the circuit design itself it would be appreciated. It's easy to over look something when operating as a lone wolf and doing design work. Give me something useful relating the actual circuit design and its potential.

The only thing I will say is that if you study the supplied diagrams, and from those draw yourself a block diagram you will find that it is not complex at all. It just has more parts since it is analog. Perhaps even post your block diagram as a aid to others. Remember this was initially designed for the purpose of seeing what level of performance is possible using modern components in an analog design, and of course as a DIY project for my own use.

I had made a video of the initial check out of the receive board I populated in my last post. It was 8 minutes long and would be helpful to who have the basic test equipment required to normally repair electronic equipment. It also showed my work space and the vacuum frame I made for holding the stencil and allowing alignment for solder paste application. I'm not going to post that video now after reading your post but will make something like it available to individuals contemplating actually building one.

For electronics technicians:

I use my HP3325A function generator and have it output a 2Hz square wave signal at 1mV and run that into a 1000:1 of attenuation to get down to about 1 micro-volt level. With the VCO idle frequency adjusted to 400Hz it should cause the VCO frequency to shift the frequency about 10Hz upwards if everything in is correct in its important TS-0 (Time Slot 0) amplification chain and its servo loop. The servo loop allows for DC coupling all the way from the front end to the VCO input if desired. On the board itself there is a 10uf capacitor into which would need to be replaced with a zero ohm resistor (1206 size). I usually order circuit boards by the dozen so if anyone is interested in building this keep that mind.

[ivconic]

I felt like I was being lectured for the second time after reading this message.

And then this: "Or is it just not working right and you're embarrassed to show it? "

Your using the wrong strategy to attempt to belittle me to get your way. First there is nothing to be embarrassed about and I'm to old at the age of 76 to play games. I'm only here to share information about my project with persons interested in building it or something like it. If my project or information is not suitable for you then I'm sure that you can find something you like since there are lots of good things on the geotech1 forum to choose from. If you wish to post something that helps me improve the circuit design itself it would be appreciated. It's easy to over look something when operating as a lone wolf and doing design work. Give me something useful relating the actual circuit design and its potential.

The only thing I will say is that if you study the supplied diagrams, and from those draw yourself a block diagram you will find that it is not complex at all. It just has more parts since it is analog. Perhaps even post your block diagram as a aid to others. Remember this was initially designed for the purpose of seeing what level of performance is possible using modern components in an analog design, and of course as a DIY project for my own use.

I had made a video of the initial check out of the receive board I populated in my last post. It was 8 minutes long and would be helpful to who have the basic test equipment required to normally repair electronic equipment. It also showed my work space and the vacuum frame I made for holding the stencil and allowing alignment for solder paste application. I'm not going to post that video now after reading your post but will make something like it available to individuals contemplating actually building one.

For electronics technicians:

I use my HP3325A function generator and have it output a 2Hz square wave signal at 1mV and run that into a 1000:1 of attenuation to get down to about 1 micro-volt level. With the VCO idle frequency adjusted to 400Hz it should cause the VCO frequency to shift the frequency about 10Hz upwards if everything in is correct in its important TS-0 (Time Slot 0) amplification chain and its servo loop. The servo loop allows for DC coupling all the way from the front end to the VCO input if desired. On the board itself there is a 10uf capacitor into which would need to be replaced with a zero ohm resistor (1206 size). I usually order circuit boards by the dozen so if anyone is interested in building this keep that mind.

You are CORRECT!
You easily saw through my childish effort to persuade you to provide a video!
But now it is clear to me, you are 20 years older than me, therefore you have much more serious experience in those stories.
So here I am publicly admitting... you saw through me!
No bad thoughts, this is a friendly forum, no complaints here.
But look at things from a different perspective...
If I was not genuinely interested in your project and if I do not like what I have seen in what you have already published;
I wouldn't even be here on the topic and I wouldn't even try to persuade you to do anything with such "provocations".
My sudden and excessive "engagement" only shows you that I like your project and that I am interested in the details.

[ivconic]

... individuals contemplating actually building one...

I was the closest to become one. Still am.
But I need video as last thing before I start.

Also, you used all the SMD, which i would like to avoid.
Bad sight, trembling hands, can't obtain all the original components from your schematics.
So I was thinking to make through-hole variant with ordinary components and available substitutions.
...
A strange phenomenon, which manifests itself in your last post... and to be fair; I notice it more and more often at myself too!
And that is: the older we are: the more stubborn we are and the easier it is to get offended.
As a younger person among two, I will give in here and say, everything is ok, no hard feelings.
But... I still need a video...